Papers
11
Total Citations
196
H-Index
6
About
Dan Bohus is a leading researcher in situated human-robot interaction, focusing on how robots can engage naturally with people in open, unconstrained environments. His work bridges spoken dialogue systems, computer vision, and social robotics to create robots that can perceive, anticipate, and adapt to human behavior in real time. Bohus is best known for his pioneering work on managing conversational engagement, introducing forecasting models that predict when a person is about to leave an interaction—a critical capability for fluid human-robot encounters. His Directions Robot, a system deployed in a real building to give directions, has become a landmark study in "in-the-wild" HRI, accumulating over 55 citations for its insights into perceptual, interaction, and output generation challenges. He has also contributed foundational concepts like "scene shaping," where robots adjust spatial formations (F-formations) to improve interaction quality, and "execution memory," enabling robots to recall shared events across engagements. With over 200 total citations across his most influential papers, Bohus’s work has shaped how researchers think about turn-taking, failure diagnosis, and sincerity in human-machine dialogue. His research continues to push the boundaries of how robots can competently collaborate with people in the messy, unpredictable real world.
Research Focus
Key Achievements
Top Papers
- 1
- 2Directions robot: in-the-wild experiences and lessons learned55 citations · 2014
- 3Dialog with robots : papers from the AAAI Fall Symposium20 citations · 2010
- 4
- 5Now, Over Here8 citations · 2020
- 6A study in scene shaping: Adjusting F-formations in the wild8 citations · 2017
- 7Turn‐Taking and Coordination in Human‐Machine Interaction6 citations · 2016
- 8
- 9Introduction to the Special Issue on Dialogue with Robots4 citations · 2011
- 10Execution memory for grounding and coordination3 citations · 2013